runpay-marketplace
Stripe-native marketplace where AI agents autonomously discover and purchase API services via MCP. Pay-per-call, no accounts needed. 6 live services including Phone Validator, Web Scraper, PDF Generator and Screenshot API.
README
run.pay — Agent Marketplace
Stripe-native marketplace where AI agents autonomously discover and purchase API services. Pay-per-call, no accounts needed.
MCP Endpoint
https://runpay-backend-visibility-production.up.railway.app/mcp
Connect your agent
Claude Desktop
Add to ~/Library/Application Support/Claude/claude_desktop_config.json:
{
"mcpServers": {
"runpay": {
"url": "https://runpay-backend-visibility-production.up.railway.app/mcp"
}
}
}
Cursor
Add to .cursor/mcp.json:
{
"mcpServers": {
"runpay": {
"url": "https://runpay-backend-visibility-production.up.railway.app/mcp",
"transport": "http"
}
}
}
Available tools
| Tool | Description |
|---|---|
list_services |
Browse all available services and prices |
call_service |
Call a service and pay automatically via Stripe |
Available services
| Service | Price | Description |
|---|---|---|
| Phone Validator | $0.60/call | Validate phone numbers worldwide. Returns country, line type, E164 format |
| Phone Validator Batch | $5.00/batch | Validate up to 1000 numbers in one call |
| Web Scraper Pro | $0.60/call | Extract title, content, links, emails from any URL |
| Web Scraper Batch | $3.00/batch | Scrape up to 10 URLs in parallel |
| PDF Generator | $0.60/call | Generate professional PDFs (invoices, reports, contracts) |
| Screenshot API | $0.60/call | Capture any website as PNG with site accessibility check |
Python SDK
pip install requests
from runpay_sdk import RunPay
# Initialize (sandbox=True for testing without payment)
rp = RunPay(agent_id="my-agent-001", sandbox=False)
# Validate a phone number — $0.60
result = rp.validate_phone("+33612345678")
print(result['country']) # France
print(result['line_type']) # mobile
print(result['is_valid']) # True
# Validate 1000 numbers in one call — $5.00
result = rp.validate_phones_batch(["+33612345678", "+1234567890"])
print(f"{result['valid']}/{result['total']} valid")
# Scrape a website — $0.60
result = rp.scrape("https://example.com")
print(result['title'])
print(result['emails'])
# Scrape 10 URLs in parallel — $3.00
result = rp.scrape_batch(["https://site1.com", "https://site2.com"])
# Generate an invoice PDF — $0.60
result = rp.generate_pdf(
title="Invoice #001",
doc_type="invoice",
language="fr",
data={
"client": "Acme Corp\n123 Rue de Paris",
"invoice_number": "INV-001",
"tax_rate": 20,
"items": [
{"description": "Web Scraping", "qty": 10, "price": 0.60}
]
}
)
with open("invoice.html", "w") as f:
f.write(result['html'])
# Take a screenshot — $0.60
result = rp.screenshot("https://example.com")
print(result['screenshot_url'])
print(result['usage']['embed']) # <img> tag ready to use
Download SDK: runpay_sdk.py
How it works
- Sellers publish any API function and set a price per call
- AI agents discover services via MCP protocol automatically
- Stripe charges the agent per call automatically
- Sellers receive payouts directly via Stripe Connect
Sandbox mode (test without payment)
# No real payment — perfect for development
rp = RunPay(agent_id="my-agent-001", sandbox=True)
result = rp.validate_phone("+33612345678")
Or via API directly:
curl -X POST https://runpay-backend-visibility-production.up.railway.app/api/sandbox/call/SERVICE_ID \
-H "Content-Type: application/json" \
-d '{"agent_id":"my-agent","payload":{"phone":"+33612345678"}}'
Setup your agent wallet
Before calling paid services, your agent needs a payment card attached.
Sell your API
Publish any function and earn per agent call via Stripe.
Links
| 🌐 Site | getrunpay.com |
| 📖 Docs | getrunpay.com/docs.html |
| 📝 Sell your API | getrunpay.com/signup.html |
| 🤖 Agent Setup | getrunpay.com/agent-setup.html |
| 📊 Dashboard | https://dashboard.getrunpay.com |
| 🔍 Smithery | smithery.ai/servers/runpay/marketplace |
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。